Uploaded September 2025 | Updated September 2026, 2 weeks ago
Follow SAI Conferences on Linkedin: linkedin.com/company/saiconference
Conference Website: saiconference.com/IntelliSys
Explore the future of text-to-image generation in this groundbreaking keynote by Prof. Cees G.M. Snoek from the University of Amsterdam at the Intelligent Systems Conference 2026. In his talk, Prof. Snoek presents NeoBabel, a revolutionary multilingual image generation foundation model that breaks free from English-centric AI and promotes global inclusivity.
🔹About the Keynote:
Traditional text-to-image generation models have primarily focused on English, limiting accessibility and reinforcing digital inequities for non-English speakers. NeoBabel addresses this challenge by natively supporting English, Chinese, Dutch, French, Hindi, and Persian, setting a new Pareto frontier in performance, efficiency, and inclusivity.
During this keynote, Prof. Snoek highlights:
★The design and development of NeoBabel, the first natively multilingual text-to-image generation model.
★A progressive training strategy that efficiently leverages limited data and compute resources.
★The creation of 124 million captioned images in multiple languages, combining real-world and synthetic data.
★A unified multimodal architecture that connects language directly to image pixels.
★Rigorous benchmarking and evaluation, including multilingual adaptations of Geneval and DPG metrics.
★Applications of NeoBabel, including multilingual inpainting, cross-cultural collaboration, and code-switching capabilities.
★Open and inclusive research approach with all data, models, and tools available for replication and improvement.
🔹About the Speaker:
Cees G.M. Snoek is a full professor of artificial intelligence at the University of Amsterdam, heading the Video & Image Sense Lab and the Human-Aligned Video AI Lab. He is also director of QUVA Lab (Qualcomm), Atlas Lab (TomTom), and OpenBots Lab (TU Delft & TNO) and serves as Chief Scientific Officer at Kepler Vision Technologies. Prof. Snoek has been a leading voice in AI, computer vision, and multilingual machine learning, driving cutting-edge research that bridges technology and accessibility.
Follow SAI Conferences on Linkedin: linkedin.com/company/saiconference
Conference Website: saiconference.com/IntelliSys
Explore the future of text-to-image generation in this groundbreaking keynote by Prof. Cees G.M. Snoek from the University of Amsterdam at the Intelligent Systems Conference 2026. In his talk, Prof. Snoek presents NeoBabel, a revolutionary multilingual image generation foundation model that breaks free from English-centric AI and promotes global inclusivity.
🔹About the Keynote:
Traditional text-to-image generation models have primarily focused on English, limiting accessibility and reinforcing digital inequities for non-English speakers. NeoBabel addresses this challenge by natively supporting English, Chinese, Dutch, French, Hindi, and Persian, setting a new Pareto frontier in performance, efficiency, and inclusivity.
During this keynote, Prof. Snoek highlights:
★The design and development of NeoBabel, the first natively multilingual text-to-image generation model.
★A progressive training strategy that efficiently leverages limited data and compute resources.
★The creation of 124 million captioned images in multiple languages, combining real-world and synthetic data.
★A unified multimodal architecture that connects language directly to image pixels.
★Rigorous benchmarking and evaluation, including multilingual adaptations of Geneval and DPG metrics.
★Applications of NeoBabel, including multilingual inpainting, cross-cultural collaboration, and code-switching capabilities.
★Open and inclusive research approach with all data, models, and tools available for replication and improvement.
🔹About the Speaker:
Cees G.M. Snoek is a full professor of artificial intelligence at the University of Amsterdam, heading the Video & Image Sense Lab and the Human-Aligned Video AI Lab. He is also director of QUVA Lab (Qualcomm), Atlas Lab (TomTom), and OpenBots Lab (TU Delft & TNO) and serves as Chief Scientific Officer at Kepler Vision Technologies. Prof. Snoek has been a leading voice in AI, computer vision, and multilingual machine learning, driving cutting-edge research that bridges technology and accessibility.
